This document provides an overview of how you can participate in improving this project or extending it. We are grateful for all your help: bug reports and fixes, code contributions, documentation or ideas. Feel free to join, we appreciate your support!!
Much of the issues, goals and ideas are tracked in the respective projects in GitHub. Please use this channel to report bugs and post ideas.
The overall hardening project is organized publicly on Trello. Feel free to join and add tasks and ideas for the overall project. https://trello.com/b/gL9v8N1q/dt-hardening
In order to contribute code please:
- Fork the project on GitHub
- Clone the project
- Add changes (and tests)
- Commit and push
- Create a merge-request
To have your code merged, see the expectations listed below.
You can find a well-written guide here.
Please follow common commit best-practices. Be explicit, have a short summary, a well-written description and references. This is especially important for the merge-request.
Some great guidelines can be found here and here.
This hardening project doesn't intend to reinvent the configuration stack for services. Aim to use official configuration projects first and provide hardening as a layer on top. The goal is remove the need for a user to configure all aspects of services and maintain security configuration. This way, the user can still configure a service using the interface provided by the official project.
- For Chef refer to the official Chef Supermarket.
- For Puppet head to the Puppet Forge and take a node of the Puppet supported modules.
These projects are generally hosted on GitHub as well.
In some cases, we in fact create the full rollout stack, but this is generally the exception (os-hardening, nginx-hardening).
- Please avoid using nonsensical property and variable names
- Use self-describing attribute names for user configuration
- In case of failures, communicate what happened and why a failure occurs to the user. Make it easy to track the code or action that produced the error. Try to catch and handle errors if possible to provide improved failure messages.
The security review of this project is done using integration tests.
Whenever you add a new security configuration, please start by writing a test that checks for this configuration. For example: If you want to set a new attribute in a configuration file, write a test that expects the value to be set first. Then implement your change.
You may add a new feature request by creating a test for whatever value you need.
All tests will be reviewed internally for their validity and overall project direction.
As code is more often read than written, please provide documentation in all projects.
Adhere to the respective guidelines for documentation:
- Chef generally documents code based explicit readme files. For code documentation please use yard-chef
- Puppet module documentation
We generally include test for coding guidelines:
- Chef follows Foodcritic
- Puppet is checked with puppet-lint
Remember: Code is generally read much more often than written.
Wherever possible, please refrain from any other formats and stick to simple markdown.